Introspection scenario

[jaromil]

We need more tools to analyse Zencode language tokens to help us develop Zencode as a concise and coherent language. For instance, stats about the most used and less used words and tokenization allow us to implement Norvig's stochastic spelling correction, as mentioned in https://github.com/dyne/Zenroom/issues/723.

The best way to implement introspection at this scope is to implement a special scenario to provide Zencode statements that will create dictionaries and other data about the information we need, having full access to a running VM, it's HEAP and STACK.